At the present time it is impossible to predict how well patients with prostatic carcinoma will respond to hormonal therapy. However, if it were possible to make this prediction, those patients who are unlikely to achieve long term benefit could be selected for other forms of treatment at an earlier time and those patients most likely to respond could be spared the risk of unnecessary radiation or chemotherapy. It is the primary goal of this proposal to develop means to predict the hormonal responsiveness of prostatic cancer. Specifically, we propose to measure steroid receptor content (androgen, estrogen, progesterone), steroid content (testosterone, dihydrotestosterone), enzymatic profiles, and morphologic characteristics of the cancer cell nucleus in malignant prostatic tissue. These findings will be correlated with quantitative aspects of response to hormonal therapy to determine whether these measurements will be of predictive value. The response to hormonal therapy will be quantitated separately for both the primary prostatic neoplasm and metastatic deposits (bone and lung). It is our hope that these correlations will prove to be of predictive value so that guidelines can be established for the proper application of hormonal therapy in the management of men with prostatic cancer. As a secondary goal, these same parameters will be correlated with the response to shemotherapy in men with relapsing prostatic cancer and the disease-free interval in men following redical prostatectomy.